Slashdot Mirror


BitTorrent Unveils Secure Chat To Counter 'NSA Dragnet Surveillance'

Hugh Pickens DOT Com writes "Jacob Kastrenakes reports on The Verge that as part a response to the NSA's wide-reaching surveillance programs, BitTorrent is unveiling a secure messaging service that will use public key encryption, forward secrecy, and a distributed hash table so that chats will be individually encrypted and won't be stored on some company's server. 'It's become increasingly clear that we need to devote hackathons, hours and resources to developing a messaging app that protects user privacy,' says Christian Averill, BitTorrent's director of communications. Because most current chat services rely on central servers to facilitate the exchange of messages, 'they're vulnerable: to hackers, to NSA dragnet surveillance sweeps.' BitTorrent chat aims to avoid those vulnerabilities through its encryption methods and decentralized infrastructure. Rather than checking in with one specific server, users of BitTorrent chat will collectively help each other figure out where to route messages to. In order to get started chatting, you'll just need to give someone else your public key — effectively your identifier. Exchanging public keys doesn't sound like the simplest way to begin a chat, but Averill says that BitTorrent hopes to make it easy enough for anyone interested. 'What we're going to do is to make sure there are options for how this is set up,' says Averill. 'This way it will appeal to the more privacy conscious consumer as well as the less technically inclined.' For now, it remains in a private testing phase that interested users can apply for access to. There's no word on when it'll be open to everyone, but with all of the recent surveillance revelations, it's easy to imagine that some people will be eager to get started."

33 of 111 comments (clear)

  1. OTR by Anonymous Coward · · Score: 2, Insightful

    How is this different from OTR?

    1. Re:OTR by heypete · · Score: 5, Informative

      How is this different from OTR?

      OTR rides on top of underlying IM protocols (e.g. AIM, ICQ, XMPP, Yahoo Messenger, etc.) and encrypts the contents of communications. IM service providers can still shut down individual accounts, monitor who is accessing them, etc., even if they cannot read the contents of messages.

      With BitTorrent Chat, the service takes advantage of the DHT (similar to "trackerless torrents" that don't have any single point of failure) to provide a decentralized, fault-tolerant means of exchanging data. There's no dependence on a single service -- all users would participate in the DHT, making it an extremely robust system.

      If I read the description properly, it's similar to "OTR-over-DHT" but there's likely substantial differences in the details.

    2. Re:OTR by Anonymous Coward · · Score: 2

      DHT doesn't seem to be robust or fault-tolerant to me.

      I also don't see how it would provide any anonymity as seems to be claimed.

      OTR over TOR might make more sense.

    3. Re:OTR by Stalks · · Score: 4, Informative

      DHT is very reliable. Once a node has been connected a while and established links with many other nodes, traffic is quick and you have the redundancy of many 100s of connections.

      Encrypting the data prior to transport and using DHT would be no worse off from TOR.

    4. Re:OTR by heypete · · Score: 2

      DHT doesn't seem to be robust or fault-tolerant to me.

      How so? The mainline DHT used for torrents has been operating without outages for years, with tens of millions of nodes taking part (with a churn of about 10 million a day). The DHT is self-healing in that if there's a small cluster of nodes that are interconnected with each other but disconnected from the main DHT, a single connection to the main DHT will result in that cluster completely rejoining the main DHT.

      I also don't see how it would provide any anonymity as seems to be claimed.

      Indeed. The system appears to provide decentralized, encrypted communications. That's a very important aspect to be sure, but it's not anonymous communication. I don't see any mention in the text or links of the summary saying that the system is intended to be anonymous, only encrypted and decentralized.

      OTR over TOR might make more sense.

      Perhaps. It'd be interesting to see a Tor-based DHT that would allow for this type of chat protocol to be extended -- that would make it both encrypted, decentralized, and anonymous.

    5. Re:OTR by GameboyRMH · · Score: 2

      Oh noes port forwarding and uPnP r so hard!!! :-(

      --
      "When information is power, privacy is freedom" - Jah-Wren Ryel
    6. Re:OTR by MacDork · · Score: 2

      Once a node has been connected a while

      Unfortunately, having a radio on for "a while" is going to put quite a lot of strain on a mobile battery. Still, it's good to see them working on this. Perhaps we could set up relays through a home server that has a power source.

    7. Re:OTR by Burz · · Score: 2

      An I2P messaging system 'I2PBote' was developed several years ago, and is based on DHT. It has the benefit of decentralization, plus the privacy/anonymity of onion routing.

      Everyone is offering an encryption mode for their apps these days. But they don't address two important factors in privacy: Revealing the who/when/where info (the metadata the NSA is so interested in, for instance), and inconsistency of use. Using a network like I2P for all/most of your communications (including email, chat, bittorrent, etc) means there is far more guesswork involved for any spy even if they want to analyze your packet timing... with all your different protocols plus forwarded traffic from other systems in the mix, it offers more privacy and anonymity than even Tor.

  2. closed source by Anonymous Coward · · Score: 2, Insightful

    You can't trust a closed source "security" app.

    1. Re:closed source by alphatel · · Score: 4, Insightful

      "It's become increasingly clear that we need to devote hackathons, hours and resources to developing a messaging app that protects user privacy"
      And should also become quite obvious that you need to start vetting coders who are infiltrating projects on behalf of the government. That good old warped 80's tinfoil hat paranoia is the only thing that will save you anymore because it seems it was never wrong.

      --
      When the foot seeks the place of the head, the line is crossed. Know your place. Keep your place. Be a shoe.
    2. Re:closed source by mwvdlee · · Score: 3, Interesting

      If the protocol is open, you could build your own app implementing it.

      --
      Slashdot social media options: AIM, ICQ, Yahoo, Jabber and Mobile Text. Why no MySpace?
  3. Cryptocat? by Anonymous Coward · · Score: 2, Funny

    Hey, guys, I got this wheel, and I don't like it. Somebody help make a new one with more corners please!

    1. Re:Cryptocat? by DuckDodgers · · Score: 5, Informative

      http://www.wired.com/threatlevel/2012/08/wired_opinion_patrick_ball/all/
      This means that in practice, CryptoCat is no more secure than Yahoo chat, ... Any host-based system that delivers the encryption engine to you each time you log in, and in which your keys reside on the server, you are never secure against the host (there’s new research on this called “host-proof hosting,” but it’s a long way from being ready to use in real applications). That means that if the host attacks you, or they fail to protect themselves, your encrypted data will be available to them. Remember that the host might attack you because someone evil has taken control of the host. If you are the hypothetical dissident in the Middle East, your government might contract a hacker to break into the CryptoCat server, Hushmail, or other host-based server, and thereby get access to all your data. Or they could bribe an employee at a host-based service. Again: in host-based security, all your security rests on your personal trust for the people at the host, and their ability to protect the server. There’s no real security in a technical sense.

  4. It's been attempted before.... by Raxxon · · Score: 2

    Started a shit-storm at AOL..... It was called WASTE...

    1. Re:It's been attempted before.... by Luckyo · · Score: 2

      Problem with WASTE is its hilarious overhead. We had a case where around min 2000s uni network's DC hub was shut down and users moved to WASTE. Suddenly all those intranet 100mbps-1gbps links that you never saw coming close to 10% usage were getting saturated as WASTE bounced every file transfer several times between nodes to obfuscate sender/receiver.

  5. Vulnerable to Social Engineering by mentil · · Score: 4, Interesting

    If the public/private key pair is created at account creation, then people accustomed to everything being in the cloud will frequently forget to backup their private key (which isn't stored on any central server). A common occurrence will be "Hey Alice, it's Bob. I lost my private key so this is my new account now." Potentially, Bob is in jail and a fed is masquerading as him.

    Also from my experience with DHT, it doesn't work unless you already know an IP running the protocol -- who you usually find through, yes, a centralized server. If that server were TOR-based it might work, but then that raises the question of what functionality is added by this protocol that a messaging program running thru TOR doesn't offer. Having Mixmaster-style message queueing in addition to onion routing would offer improved resistance to topology attacks as well. I'm referring to TOR's hidden services protocol, by the way, rather than the standard web proxy where an unencrypted message would be sent to a messaging server after several encrypted hops.

    --
    Corruption is convincing someone that the selfless ideal is the same as their selfish ideal.
    1. Re:Vulnerable to Social Engineering by MoonFog · · Score: 5, Interesting

      So what you're saying is that the police forces would have to use old-fashioned police work and target individuals they suspect rather than mass collect everything? Yes, clearly that proves this is a wasted effort.

    2. Re:Vulnerable to Social Engineering by Joce640k · · Score: 5, Insightful

      This is to stop mass spying/trawling.

      If your contacts are all in jail then you have bigger problems.

      --
      No sig today...
    3. Re:Vulnerable to Social Engineering by Anonymous Coward · · Score: 4, Interesting

      yes but they could not do it to everyone, only people who they had good grounds to obtain a warrant, even without the need for a warrant they would still have not acquired the full message history from everyone all the time as this would have been too much work. All this does it put this situation back how it was.

    4. Re:Vulnerable to Social Engineering by RabidReindeer · · Score: 2

      I hate to break the news to you , but "old-fashioned police work" also involved wire tapping and intercepting post. Its whats known as gaining evidence. Now in a perfect world it wouldn't be necessary , but unfortunately we don't live in a perfect world and a lot of people don't play by the rules and oddly enough they usually say "no" if someone asks if they've commited a crime. Therefor the authorities have to get information on their activities, a lot of which comes from correspondence in whatever form it takes.

      Hope that clears things up for you since you're clearly new to this reality.

      Whoosh!

      target individuals they suspect rather than mass collect everything

      Nothing was said about no longer doing wire tapping etc. Merely that the mere act of existence should no longer make you a suspect on equal footing with the senior counsel of Al-Qaeda.

  6. Re:honey by Anonymous Coward · · Score: 3, Insightful

    This attitude needs to go away. Now. I have EVERYTHING to hide.

  7. I'm just thinking out loud here.. by Arith · · Score: 2, Interesting

    I'm not exactly a crypto-guru.. but if exchanging a key with your friend to ESTABLISH secure chat.. wouldn't you first have to send that key through unencrypted channels? - assuming you are far enough away that face-to-face isn't an option (and in that case, why even use this?)

    1. Re:I'm just thinking out loud here.. by Joce640k · · Score: 3, Informative

      The thing about PUBLIC key cryptography is that the encryption keys are PUBLIC, it doesn't matter who knows them.

      --
      No sig today...
    2. Re:I'm just thinking out loud here.. by Anonymous Coward · · Score: 5, Informative

      This explains it very simply.

      You can exchange a piece of information without exposing the full picture to a 3rd party.

    3. Re:I'm just thinking out loud here.. by root_brewski · · Score: 2

      Diffie-Hellman is a commonly taught example of how to do this. The video linked above explains it well, but here's a wiki link if you want to read text instead.

  8. Re:It could be a honey pot by Joce640k · · Score: 2

    when I'm using something that supposed to be secured, I'm still taking extra precautions.

    At the very least, every single connection that I'm making I scramble my MAC address.

    And btw, the other day there was a news about someone got apprehended despite hiding behind a TOR - you just never know when they will come at you, and when they want to get you, they can.

    If you're doing things that make them specifically target you then they'll probably get you anyway (and that's as it should be).

    This new is to stop mass spying/recording of every conversation (which is what they're doing now).

    --
    No sig today...
  9. user acceptance? by Tom · · Score: 2

    The primary issue will be the same as for PGP (anyone use that? wait, let me rephrase that: anyone know of any non-geek people who use it?): User acceptance.

    Unless it's as easy not only to use but also to add contacts as FB chat, AIM, ICQ, Skype, Google+, etc. etc. it won't get the critical mass it needs.

    What good is a secure chat if you don't have anyone you can chat to?

    --
    Assorted stuff I do sometimes: Lemuria.org
  10. These types of things get a bad rep by Daniel+Hoffmann · · Score: 2

    Everyone should have their communications privacy by default, not having to hunt downs means to keep their privacy. Services like these end up being used mostly by paranoids and people with malicious intent. So in some respects the government officials have a point in wanting to shut down initiatives like these. On the other hand initiatives like these only exist because the government wants to control everything.

    1. Re:These types of things get a bad rep by mwvdlee · · Score: 4, Funny

      I use these services because I have nothing to hide and like our caring government.
      Since I have nothing to hide, my caring government would be wasting resources trying to monitor and read my communication.
      By making it impossible for them to do so, I'm saving them effort, time and money.

      --
      Slashdot social media options: AIM, ICQ, Yahoo, Jabber and Mobile Text. Why no MySpace?
  11. Re:bedroom spy camera by KingOfBLASH · · Score: 3, Insightful

    Breaking into houses takes time and resources. Much more so then if the NSA can simply watch all your chat, archive it in a huge data center in utah, and then do a search through your histories.

    So when spy agencies have to work the old fashioned way, even if they disrespect the need for warrants they still can't spy on everybody. If they can just do a search through your data, they are effectively watching everyone

  12. PATENTS and veiled threats at Open Source by FriendlyLurker · · Score: 2

    How many different BitTorrent clients are there? How did that happen...?

    It would appear that Bittorrent the company considers the healthy bittorrent/client ecosystem to be a mistake not to be repeated. Like this chat protocol, they have also announced a P2P Streaming protcol - their implimentation will be closed source encumbered with patents that they have threated to use against anyone wishing to start an alternative open client. So even when they openly publish the protocol, it is still of no use the open source community. Don't believe me, take this quote from the horses mouth:

    “We want people to use and adopt BitTorrent Live. But we aren’t planning on encouraging alternative implementation because [Insert pathetic excuse here]. We want to ensure a quality experience for all and this is the best approach for us to [i.e more pathetic excuses to close source the system],” Cohen told TorrentFreak.

    So, yeah, You can read the protocol spec but try to impliment it and we will "discourage" you - i.e. use out patent(s) to clobber your OS project to oblivion. Personally I hope the open source community can take these interesting initiatives, design around the patents and make a true P2P Streaming and secure chat system ecosystem - because it appears that Bittorrent the company has fallen far from its early success of kicking off the truly open bittorrent protocol, sadly.

  13. Another proprietary mess - a pity. by RanceJustice · · Score: 2

    Much like MEGA, the other projects of BitTorrent labs (most notably - Snyc), and a whole host of pseudo-security minded programs and services popping up recently, this is sadly proprietary bullshit. Much like BitTorrent Inc absorbing uTorrent as the main client etc... they've repeatedly demonstrated that they view their greatest success - the Bit Torrent protocol itself, as a mistake to be avoided. Why did BitTorrent itself grow to be so prevalent? Exactly the thing they seem to hate - its openness. BitTorrent protocol and most of its extensions (ie DHT, uTP, PEX and more) are all free and open source, to be implemented in a variety of clients. This is its greatest strength, from the slashdot-reading hacktivist running Deluge/Transmission/rTorrent, to World of Warcraft's client updater/patcher, BitTorrent is not just a great protocol for both tracker-based and trackerless sharing, but its implementations are as wide as can be and interoperable.

    I am not sure why BitTorrent Inc has decided to treat this as a weakness, and develop yet another proprietary software-as-a-service, centrally managed debacle. While there seems to be some casual lip service paid to FOSS and promises of openness, I haven't seen any examples that they're actually interested in such things. For instance, the javascript Torque API which is supposed to bring BitTorrent to the web browser, doesn't seem to be compatible with any clients aside from the official BitTorrent/uTorrent clients themselves! Other "labs" projects like Live, Surf, and Sync are similar in this regard, being designed only for approved first-party clients.

    So long as this ideal reigns, I won't be using these projects. Especially when it comes to privacy and security it is simply too important than to trust a proprietary, unverifiable item of this sort. There are already a variety of projects that offer better privacy and more secure messaging - RetroShare for instance. If you're interested in some of the best, check out www.prism-break.org for a directory of privacy and security respecting, mostly FOSS, programs for many uses. Until those like BitTorrent Inc wake up and realize that openness is one of their greatest strengths, I don't see any reason to consider what they provide.

  14. Can't do anonymous + decentralized + enc-envelope by aztracker1 · · Score: 2

    You could allow something like this to be anonymous, by allowing addressing to be a relatively poor hash that a message to you would match many people and from there signature checks could determin that the message was in fact for you. The problem comes in that any such system is open to channel poisoning.. many fake messages to a block so that the decryption time becomes costly, and/or you get so many garbage messages that you can't filter out the "real" stuff. This just shows you can do encrypted + decentralized, but not really anonymous. Any solution I've thought through has to give up one of anonymous, decentralized or full-encryption of the message envelope

    --
    Michael J. Ryan - tracker1.info